Anomaly and novel events detection is vital to the financial services industry. These events may often be indicative of illegal activities such as fraud, risk, identity theft, network intrusion, account takeover and money laundering which may result in undesired outcomes such as disruption in service and other breakdowns. As financial environments change, digital adoption grows and the data moves at increasing speed and volume, the problem of detecting anomalies in real time at large scale becomes increasingly challenging. This is further compounded by the fact that more and more anomaly detection applications require operational decision making in real time. Several new ideas are emerging to tackle this challenge, including semi-supervised learning methods, deep learning based approaches and network/graph based solutions. These approaches must often be able to work in real time by consuming and processing large volumes of data produced in real time.
Multiple industries are witnessing an exponential increase in the availability of streaming large volume of data. This is certainly true for the financial industry. Largely driven by an increase in instrumentation both on the front end and back end applications, we now have enormous number of systems that produce continuously changing data in real time. This data is representative of the health and well-being of the system and applications. Hence, any deviations from the past behavior that is unusual and significant is of special interest and may require action. Our streaming dataops product for financial services delivers critical real time streaming data intelligence.